基于UKF的光纤惯导误差在线估计

On-line FINS Error Estimation Based on UKF

为利用星光观测在线估计弹载光纤陀螺捷联惯导(FSINS)的陀螺零偏、标度因数和安装误差,并依据估计结果对导航参数进行补偿,设计了基于旋转弹体的FSINS误差分离方法,推导了误差估计的状态方程和观测方程,设计了UKF滤波器,并在星光姿态确定的不同动态误差条件下进行了仿真,结果显示该方法能较好的估计陀螺零偏常值漂移、标度因数误差和安装误差,表明了方法的正确性,也说明了误差估计效果与星敏感器姿态确定精度的关系。

In order to on-line estimate the constant zero bias,scale factor and fixing errors of fiber optic gyro strapdown inertial navigation system（FSINS） applied in missiles and correcting navigation parameters measured by FSINS,a FSINS error separation algorithm based on rotating missile body was designed,and the state equations and observation equations were deduced based on unscented Kalman filter（UKF）.The simulation was made to check the UKF in different dynamic error,and the result shows the method can estimate the gyros＇ zero bias,scale factor errors and fixing error effectively and shows the relation between estimation precision and CNS attitude determined precision.